« Hal Varian on Short Selling | Main | links for 2008-10-05 »

Saturday, October 04, 2008

What Caused the Financial Crisis?

An article in the NY Times, "Pressured to Take on Risk, Fannie Hit a Tipping Point," is causing many people to wonder if Fannie and Freddie caused the financial crisis.

First, let me clarify the question. We are asking what caused the housing bubble, and, by definition, the cause cannot be explained by changes in an underlying market fundamental. I don't mean that we can't point to, say, a rumor that led to a rapid increase in the price of some good as speculators rush in, just that bubbles - by definition - are divorced from market fundamentals.

I think a more interesting question is what sets the stage for a bubble to emerge - what allows the rumor, irrational exuberance, etc., to express itself as a bubble? One thing that is needed is liquidity and credit, some way of substantially increasing demand. This is the air that inflates the bubble. Even if all the other conditions for a bubble to emerge are present, if there is no way to inflate the bubble - no way for speculators to rush in and drive up the price - then it won't inflate.

We already know that there was enough available liquidity to inflate a housing bubble. So something went wrong in these markets that allowed the bubble to emerge and then pop, and this is causing us immense problems right now, but what was it?

I think the most important factors are agency problems, the mis-pricing of risk, and the failure of securitization to distribute risks across the financial system.

With respect to the agency issues, there is a long chain between the home buyer, the mortgage broker, and, ultimately, the sliced and diced complex securities that nobody fully understands. Let's take one step in the chain, that of a a bank or mortgage broker, either one. Suppose they are paid a fee, i.e. by the number of mortgages that pass through their hands each month (as, essentially, they were). The more mortgages they can push through, the higher their income. They are required to meet certain guidelines as they do this, but so long as their income depends upon the number of mortgages passing through their hands and not what happens to the mortgages later on - so long as it is a fee-based system - they have every incentive to push the guidelines as hard as they can and to find a way around them whenever possible.

If mortgage brokers had done their job and only made loans to people who could pay them back (i.e. with "reasonable" levels of default), we wouldn't have a financial crisis. So right away, in nearly the first step of the chain, we have to ask what went wrong, why they were willing to take so many questionable loans. The problem is what economists call an agency issue. The brokers had no stake in the outcome once the mortgages left their hands. The same with banks, all they had to do was process the mortgages, package them up, then sell them and collect their fee.

Think about the incentives here. Suppose you are a mortgage broker and you begin to suspect that the bubble will pop soon, that all this lucrative business might end. To protect the business, should you get worried and start checking mortgages more carefully to make sure that things don't get further out of hand? No, you should accelerate what you are doing, write even more mortgages - nothing you can do can stop the bubble from popping, you are just one of many, many brokers far down the chain - so why not collect as many fees as possible before the gravy train ends? What if everyone thinks this way, and they all rush to sell as many of these things as they can? Mania.

A solution to this is to give each person in the chain a stake in the future outcome of the mortgage. If mortgage brokers' income had been connected to a financial instrument that pays off according to the future performance of the mortgages they write, would they have behaved differently? Probably. (What about homeowners, why didn't they say no? Don't they have a stake in the future price of the home? Homeowners in non-recourse states - and more generally - were basically granted cheap options on their homes. The downside was protected and they had no reason to effectively monitor risk. If prices fell, they could just walk away and know that their other assets remained safe and that their credit reputations could be restored with time. Of course, if everyone walks away other assets such as retirement savings don't remain safe, but that doesn't change the incentive on an individual level.)

Ah, you say, but as you go up the chain why didn't people refuse to take the financial paper, why didn't they conclude it was too risky? The risky mortgages don't have to be stopped at the bottom, this is a linked chain, so why weren't they stopped higher up in the chain where the stakes are higher? Isn't that where Fannie, Freddie, and moral hazard rear their ugly heads? Did they encourage and allow this risky paper to pass through the system?

The mis-pricing and mal-distribution of risk played a key role here (along with poor management decisions in cases where alarms were raised). The agency issues above, and the consequences of the failure to predict and distribute risk are much more important than any moral hazard issues arising from the implicit government guarantee granted to Fannie and Freddie.

Institutions in the shadow banking sector were willing to take large volumes of risky loans as they came up through the system. Why?

The people at the top of this complex chain did not fully understand the risks the were assuming when they took on the subprime business, or, rather, when they took on the complex securities derived from the subprime business. When the bubble popped, it shouldn't have been a big problem if the risk assessment models they relied upon had been correct, and if securitization had distributed the risk as promised. As Brad DeLong notes:

  • There is $11T if U.S. mortgages
  • There is $60T of global financial assets
  • Even if we had $2T of losses on mortgage-backed securities that shouldn't pose a big problem for Wall Street--actually 48th and Park Avenue

So if the risks had been distributed fairly evenly, it's much less likely that we'd be in this mess (the losses of 2T - an intentionally high-balled number - are only 1/30th of global financial assets). It wasn't the misprediction of the level of risk that was the biggest problem, the losses could have been absorbed, it was the (unintended) concentration of risk through the failure of securitization that was the most problematic.

Fundamentally, then, it was the agency problems and the failure of risk prediction and distribution models that allowed the bubble to inflate and then cause big problems after it popped. But back to Fannie and Freddie. The willingness of the non-traditional banking sector - the shadow banking system - to take on these risky assets and still pay investors a relatively high return put tremendous pressure on Fannie and Freddie to follow suit. And their response was unwise - Fannie and Freddie followed the shadow banking sector downward. There is lots to fault in the behavior of Fannie and Freddie and in government oversight of them - the decisions of management, the lobbying efforts that were funded by their ability to extract a premium from the implicit government guarantee - all of this was a big problem. The bubble, and later the financial crisis expressed itself in these institutions, and they may have also contributed to it to some extent as they took on more risky securities when their business began to go elsewhere. But the agency issues and the failures of risk models and securitization would have created problems in the largely unregulated shadow banking sector even if these two institutions had taken on nothing but the safest of mortgages. The bubble still would have inflated in the shadow banking system - maybe it's a little smaller, I don't know - but it still would have been large enough to cause big problems when it burst. The best behavior of Fannie and Freddie would not have been enough to stop the bubble from inflating in other parts of the financial sector, and then turning into a full fledged financial crisis as housing prices plunged.

The problems we are having were caused when lots of available liquidity rushed past the checks and balances that proper agency provides in pursuit of promises that risk models and complex securities did not deliver. The unexpected losses alone might not have caused a crisis had the losses been widely distributed, but, the losses were concentrated and hidden in ways that created widespread fear and threatened the entire system. Getting rid of that fear is not going to be easy.

[Update: Given some of the responses elsewhere to this post and others like it, let me add one more thing. Asking the question "what caused the financial crisis," thinking about it, and then arguing that Fannie and Freddie were not the primary driving forces behind the financial meltdown (though they could have affected the size of the problem as noted above) is not the same as defending Fannie and Freddie. Whether are not Fannie and Freddie are performing a useful function, and if they are performing a useful function how they should be structured going forward is not a question I've fully resolved. The market failure they are addressing is not entirely evident to me, and until I understand how they improve the efficiency of these markets, I won't take a position. They certainly should not operate as private entities with an implicit government guarantee as before - that's what set up the situation where the implicit guarantee could be exploited profitably and used to fund lobbyists and ad campaigns to make sure the golden goose kept laying eggs. However, I have posted arguments from other people arguing for their existence, and I am thinking about those as well as arguments against their continuation. In any case, something to guard against, I think, is to inappropriately blame Fannie and Freddie for the financial crisis and then use that as a reason to shut them down irrespective of any useful function they might serve. So when I see those with an agenda against government intervention trying to do just that - arguing honestly in some cases and dishonestly in others that Fannie and Freddie were a big factor in the crisis so they can use them as an example of government intervention gone awry and also shut them down - a double bonus in their eyes - I have tried to present evidence and arguments that the cause lies elsewhere. But as I said, that is not the same as defending their existence. My interest is in understanding the true cause of the financial crisis and in stopping it from happening again - and to avoid getting stuck on wrong arguments along the way - not in using the crisis to argue about whether Fannie and Freddie ought to continue as government supported institutions. That can wait for another day.]

    Posted by on Saturday, October 4, 2008 at 04:05 PM in Economics, Financial System, Housing, Market Failure | Permalink  TrackBack (1)  Comments (105)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference What Caused the Financial Crisis?:

    » The Lamps Are Going Out from Econlog

    No major newspaper would run any op-ed from me. I gave up and sent one to American.com, which you can see here. The bottom line: The financial bailout isn't as bad as Main Street thinks. It's worse. I think... [Read More]

    Tracked on Saturday, October 04, 2008 at 06:24 PM


    Feed You can follow this conversation by subscribing to the comment feed for this post.